FTBFS should not be dependent on lack of technical knowledge of a builder or lack of detail documentation. Perhaps it should be an aggregation of results from multiple attempts from multiple independent builders.
At nostr:npub1j9kttlc86w63emmldd4h74rekyqpksqup6p9trhp5gjsf374qlyszvuswx we have a difference about when to mark a wallet as "failed to build from source" aka "ftbfs".
The one position is to mark products as "ftbfs" if after spending some hour or two the build still fails.
The other position is that maybe with more effort it would not fail, so marking it "ftbfs" is more a sign of lack of knowledge and determination than a failure on part of the wallet provider.
What do you think? When should a wallet be "failed to build from source"?
Discussion
If only more people would care to try to build. Sadly most blame themselves when they fail and don't even dare to open an issue to resolve what they couldn't figure out.
I was in charge of an Android Bitcoin wallet for several years. My stance is:
* The bar is "reasonably technical person"
* I'm above that bar
* If I fail, a negative verdict is warranted
There is environment specific technical ability. One might be good at troubleshooting C builds and not Java. Perhaps a standard environment should be used, like a VM image, where many agree it's sufficiently configured for building a specific source language. You might find wallet devs aim to get it to build on that standard environment.
How to aggregate a bunch of 👍 and 👎 into a verdict?
I would love to open up to multiple reviewers suggesting a verdict but currently I'm even behind myself in terms of validating builds and have little hope to get more people to build wallets. But even if people did, what would I make of it? The provider might have some sock puppet accounts that just claim whatever. I have little hope to get the most trusted users to review products that are not their own.
